Small Ribosomal Protein Subunit S7 Suppresses Ovarian Tumorigenesis through Regulation of the PI3K/AKT and MAPK Pathways

Ziliang Wang1☯, Jing Hou2☯, Lili Lu4☯, Zihao Qi1, Jianmin Sun3, Wen Gao3, Jiao Meng1, Yan Wang1, Huizhen Sun1, Hongyu Gu1, Yuhu Xin1, Xiaomao Guo2*, Gong Yang1*

1 Cancer Institute, Fudan University Shanghai Cancer Center, Department of Oncology, Shanghai Medical College, Fudan University, Shanghai, China, 2 Department of Radiation Oncology, Fudan University Shanghai Cancer Center, Department of Oncology, Shanghai Medical College, Fudan University, Shanghai, China, 3 Department of Gynecological Oncology, Fudan University Shanghai Cancer Center, Department of Oncology, Shanghai Medical College, Fudan University, Shanghai, China, 4 Life and Environment Science College, Shanghai Normal University, Shanghai, China

Abstract

Small ribosomal protein subunit S7 (RPS7) has been reported to be associated with various malignancies, but the role of RPS7 in ovarian cancer remains unclear. In this study, we found that silencing of RPS7 by a specific shRNA promoted ovarian cancer cell proliferation, accelerated cell cycle progression, and slightly reduced cell apoptosis and response to cisplatin treatment. Knockdown of RPS7 resulted in increased expression of P85α, P110α, and AKT2. Although the basal levels of ERK1/2, MEK1/2, and P38 were inconsistently altered in ovarian cancer cells, the phosphorylated forms of MEK1/2 (Ser217/221), ERK1/2 (Thr202/Tyr204), JNK1/2 (Thr183/Tyr185), and P38 (Thr180/ Tyr182) were consistently reduced after RPS7 was silenced. Both the in vitro anchorage-independent colony formation and in vivo animal tumor formation capability of cells were enhanced after RPS7 was depleted. We also showed that silencing of RPS7 enhanced ovarian cancer cell migration and invasion. In sum, our results suggest that RPS7 suppresses ovarian tumorigenesis and metastasis through PI3K/AKT and MAPK signal pathways. Thus, RPS7 may be used as a potential marker for diagnosis and treatment of ovarian cancer.

Citation: Wang Z, Hou J, Lu L, Qi Z, Sun J, et al. (2013) Small Ribosomal Protein Subunit S7 Suppresses Ovarian Tumorigenesis through Regulation of the PI3K/AKT and MAPK Pathways. PLoS ONE 8(11): e79117. doi:10.1371/journal.pone.0079117 Editor: Salvatore Papa, Institute of Hepatology - Birkbeck, University of London, United Kingdom Received May 28, 2013; Accepted September 18, 2013; Published November 11, 2013 Copyright: © 2013 Wang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by National Nature Science Foundation of China (#91129721) for Gong Yang and Young Backbone Foundation of Fudan University (11L-38B) for Ziliang Wang. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. * Email: [email protected] (XMG); [email protected] (GY) ☯ These authors contributed equally to this work.

Introduction carcinoma 20 years ago [6], RPs have been reported to have important roles in many cancers. For example, silencing of Ovarian cancer is the most lethal gynecological malignancy ribosomal proteins L26 and L29 by small interfering RNAs worldwide [1]. Ribosomal proteins (RPs) play essential roles in (siRNAs) inhibited the proliferation of human pancreatic cancer the formation of a fully functional ribosome, which is cells [7]. Knockdown of the ribosomal protein RPL19 by siRNA responsible for protein synthesis in both prokaryotic and abrogated the aggressive phenotype of human prostate cancer eukaryotic cells. Over the past decades, more and more RPs [8]. The ribosomal protein L6 promoted cell cycle progression have been identified to exert diverse functions in addition to through up-regulating cyclin E in gastric cancer cells [9]. protein synthesis [2,3]. Some studies have shown that the Overexpression of the ribosomal protein L15 was associated dysfunction of ribosomal proteins results in dysregulation of with cell proliferation in esophageal cancer [10]. protein translation, leading to the development of many Overexpression of RPL36a enhanced cellular proliferation in cancers [4,5]. However, the underlying mechanisms are largely hepatocellular carcinoma [11]. Another study demonstrated unknown. that RPS7 may block the function of p53/MDM2 and abrogate Some studies have shown that RP genes are oncogenes in the Ling Zhi-8-induced lung cancer cell proliferative advantage human tissues. Ever since RPS19 was found to have a close [12]. In our previous work, we used zebrafish as a model and relationship with progression and differentiation of colon found that at the early stage of embryonic development, RPS7

PLOS ONE | www.plosone.org 1 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

suppressed cell apoptosis and cell cycle progression through Retroviruses expressing RPS7 shRNA or GFP shRNA were dysregulation of p53 [13,14], indicating that RPS7 might be an produced by transfection of pBabe/U6/shRPS7 or pBabe/U6/ important factor in human cancers. However, in 2007, RPS7 shGFP into phoenix amphotropic cells and used to infect target was also identified to activate p53 and MDM2, which cells (SKOV3 , OVCA429, and OVCA433) by using a method subsequently promoted cellular apoptosis and inhibited cell described before [19]. Briefly, cells were infected twice for a proliferation [15]. Thus, the regulatory function of RPs in total of 6 days (3 days for each infection) and the positive tumorigenesis is controversial and still needs further clones were selected with puromycin (200 ng/mL) for 10–14 investigations. days to establish stable cell lines expressing shRPS7 or To elucidate the biological function of RPS7 in ovarian shGFP. The resulting cells were used for following experiments cancer, we silenced the expression of RPS7 by specific shRNA without addition of puromycin. in ovarian cancer cell lines and tested the effects of RPS7 on cell proliferation, apoptosis, cell cycle, migration, invasion, and Cell proliferation tumorigenesis. Our results demonstrate that RPS7 suppresses Cells were detached by using trypsinization and washed ovarian tumorigenesis and metastasis through dysregulation of twice with PBS. 2×103 cells per well were seeded in 96-well PI3K/AKT and MAPK signal pathways. culture plates (Corning Inc., Corning, NY) in 100 μl medium and cultured for 2, 4, 6, 8, 10 days. Cell growth was detected Materials and Methods using Cell Counting Kit-8(CCK-8) (Dojindo Laboratories, Kumamoto, Japan) according to the manufacturer’s Ethics statement instructions. Absorbance at 450 nm was measured with a All mouse experiments were approved by the Institutional microplate reader. The assay was independently repeated Animal Care and Use Committee of Fudan University Shanghai three times. Cancer Center, and performed following the institutional guidelines. Cell invasion and migration To test cell invasion, we used a high throughput screening Cell lines and cell culture multi-well insert 24-well two-chamber plate (BD Biosciences, Human epithelial ovarian cancer cell lines OVCA433, San Jose, CA), with an 8-µm (pore size) polycarbonate filter OVCA429, OVCA420, SKOV3, and retroviral packaging cells between chambers. 2.5 × 103 cells of SKOV3-shRSP7, (Phoenix amphotropic cells) were purchased from American OVCA429-shRPS7 or OVCA433-shRSP7 and their Type Culture Collection (Manassas, VA). Immortalized human corresponding controls expressing shGFP were added in upper ovarian surface epithelial cell line T29 was derived from the chamber and allowed to invade at 37°C for 24 hours toward a normal ovarian surface epithelial cell line IOSE29, which was lower reservoir containing medium plus fibronectin (20 µg/mL). described elsewhere [16,17]. OVCA433, OVCA429, T29 and The cells were then fixed in 100% methanol for 30 minutes and phoenix cells were maintained in DMEM medium stained with Giemsa solution for 10 minutes. The invasive cells supplemented with 10% fetal bovine serum, 2 mM l-glutamine, were counted as those passed through the membrane nonessential amino acids (1%), 1 mM sodium pyruvate, separating the chamber. All cells were counted at ×200 penicillin (100 units/mL), and streptomycin (100 µg/mL). magnification under a microscope. The assay was repeated SKOV3 and OVCA420 cells were maintained in RPMI 1640 three times with duplicate. medium containing 10% fetal bovine serum, 2 mM l-glutamine, To examine cell migration, cells were incubated in 6-well penicillin (100 units/mL), and streptomycin (100 µg/mL). plate over-night to yield monolayer confluence for scratch assay. Scratches were made using a pipette tip and Cisplatin treatment photographed immediately (time 0) and 24 hours’ later. The Cisplatin was purchased from QiLu pharmaceutical company distance migrated by the cell monolayer to close the scratch (Shanghai, China). Stock concentration of cisplatin was 5mg/ml area during the time period was measured. Results were and the concentration used to treat ovarian cancer lines was analyzed as migration index, which was the ratio of the cell 2.5 µg/mL. The apoptosis of cells was detected by Nicoletti migration distance at 24h to that at 0h.The assay was carried assay [18]. out in triplicate and repeated three times.

Generation and retroviral delivery of small hairpin RNA Anchorage-independent colony formation (shRNA) against RPS7 mRNA To detect anchorage-independent colony growth, soft agar The DNA oligonucleotides used to generate shRNA against assay was performed with cells stably expressing either the open reading frame of RPS7 mRNA (positions 167–189) shRPS7 or shGFP according to the previously published were 5-TCGGAAAGCTATCATAATCTTTa-3. pBabe/U6/ method [19,20]. Briefly, 5 × 104 cells were suspended in 2 mL shPRS7 was generated according to the previously reported of medium with 0.35% agarose (Life Technologies), and the method [19,20]. The control vector was similarly constructed by suspension was placed on top of 5 mL of solidified 0.7% directly inserting oligonucleotides encoding small hairpin RNA agarose. Triplicate cultures of each cell type were maintained against green fluorescence protein mRNA (shGFP) into for 14-28 days at 37°C in a 5% CO2 atmosphere, and fresh pBabe/U6/puromycin [19,21]. medium was fed every 7 days. The number of colonies > 50

PLOS ONE | www.plosone.org 2 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

µm (~100 cells) in diameter in each dish was counted at 14 to against the following proteins were obtained from Santa Cruz 20 days. The assay was repeated three times in duplicate. Biotechnology: RPS7, BAX, Bcl-2, cyclin-dependent kinase (CDK) 2, CDK4, cyclin B1, cyclin D1, p21waf1/cip1, p27cip/kip, Cell cycle and apoptosis MMP2, MMP9, MMP13, AKT1, AKT2 and E-cadherin. Cells (1–2 × 106) were fixed with 4 mL of cold 75% ethanol at Antibodies against the following proteins were from Cell 4°C for a minimum of 4 hours and stained with 200 µL of Signaling Technology (Danvers, MA): β-actin, BAD, Bcl-XS, β- propidium iodide (50 µL/mL; Sigma-Aldrich) and 20 µL of cantenin, MEK1/2, ERK1/2, JNK1/2, P38, P85α, P110α, RNase (1 mg/mL; Sigma-Aldrich) in a 37°C water bath for 15 to pMEK1/2 (Ser217/221), pERK1/2 (Thr202/Tyr204), 20 minutes. Cell cycles were determined by FACStation (BD pJNK1/2(Thr183/Tyr185) and pP38 (Thr180/Tyr182). The Biosciences) and analyzed by using CellQuest software and a secondary antibodies were F(ab)2 fragment of donkey anti- published method [19]. The assay was repeated three times in mouse immunoglobulin (product NA931) or of donkey anti- duplicate. This method was also used to detect the percentage rabbit immunoglobulin (product NA9340) linked to horseradish of sub-G1 cells, which represents cellular apoptosis after cells peroxidase from Amersham Biosciences (Little Chalfont, were treated with or without cisplatin. Buckinghamshire,UK). Immunoblotting reagents were from an To detect apoptosis, 1 × 105 cells were stained with Annexin electrochemiluminescence kit (Amersham Biosciences). V and propidium iodide according to the instruction of the Annexin V–fluorescence apoptosis detection kit I (BD Xenograft tumors in nude mice Biosciences PharMingen), and to analysis with a FAC Station All mouse experiments were performed following the equipped with CellQuest software. The percentage of apoptotic institutional guidelines approved by the Institutional Animal cells was calculated in terms of peaks (M2) in the histogram, Care and Use Committee of Fudan University Shanghai representing an early apoptotic population (Annexin V+/PI−) Cancer Center. Xenograft tumors were either subcutaneously among the total cells analyzed. The experiment was done in or intraperitoneally generated to monitor tumor growth or duplicate and repeated three times. metastasis of SKOV3-shRPS7 and OVCA433-shRPS cells and their controls. 4- to 6-week-old BALB/c athymic nude mice Real-time PCR (Department of Laboratory Animal, Fudan University) were Total RNA from 2 × 106 cells for each cell line was isolated used for the experiment and kept in a pathogen-free by using Trizol reagent (Invitrogen, Carlsbad, CA). All RNAs environment. For the subcutaneous injection group, 5 × 106 were then reversely transcribed into cDNAs that were suitable cells were used for each injection, and each mouse received for real-time PCR analysis using the ExScript RT-PCR kit two injections in bilateral flanks. Total five mice were used for (TaKaRa, Japan). To synthesize cDNA, 0.5 mM each cell line. The date on which the first grossly visible tumor deoxynucleoside triphosphate, 50 pmol random hexamers, 50 appeared for subcutaneous injection was recorded, and the U ExScript reverse transcriptase (200 U/μl), 10 U RNase tumor size was measured every 3 days. Two-dimensional inhibitor, 500 ng total RNA, and 1 × reaction buffer were mixed measurements were taken with an electronic caliper, and tumor in each reaction tube (10μl per reaction) and then incubated at volume was calculated with the use of the following formula: 42°C for 15 min, followed by a 2-min incubation at 95°C to tumor volume (in mm3) = a × b2 × 0.52 [23], where a is the inactivate the ExScript reverse transcriptase. Oligonucleotide longest diameter, b is the shortest diameter, and 0.52 is a primers for RPS7 were 5’-GTCGTCTTTATCGCTCAGAG-3’ constant to calculate the volume of an ellipsoid. When a tumor (Forward primer) and 5’-TGTCAGAGTACGGCTCCTG-3’ reached 1.0 cm in diameter, the mouse was killed by exposure (Reverse primers). Oligonucleotide primers for GAPDH were to 5% carbon monoxide. Four tumors per cell line were 5’- GGCCTCCAAGGAGTAAGACC-3’ (forward primer) and 5’- excised, fixed in 10% formalin overnight, and subjected to CAAGGGGTCTACATGGCAAC-3’ (reverse primers). All routine histological examination by investigators who were amplifications and detections were carried out in the Applied blinded to the tumor status. Biosystems Prism 7900 system (Applied Biosystems, Foster For the intraperitoneal injection group, eight mice were used City, CA) using the ExScript Sybr green QPCR kit (TaKaRa) for each cell line and each mouse received one injection of 5 × and the following program: 95°C for 10 s, one cycle; 95°C for 5 106 cells. Mice were observed for lethargy, poor appetite, and s, 62°C for 31 s, 40 cycles; followed by a 30-min melting curve abdominal enlargement, and sacrificed timely before natural collection to verify the primer dimers. Statistical analysis was death occurred. Tumor nodules were counted, weighed, and performed using the 2-△△CT relative quantification method. measured for their numbers, weights, and volumes. Animal assays were repeated twice. Immunoblotting analysis To analyze RPS7 expression in cells, we prepared cell Immunohistochemical staining and analysis lysates at 75% of confluence using 500 µL of Samples from 20 xenograft mouse tumors were used for radioimmunoprecipitation assay buffer (RIPA, 25 mM Tris–HCl immunohistochemical staining of E-cadherin and β-catenin at pH 7.6, 150 mM NaCl, 1% Nonidet P-40, 1% sodium expression. Antibodies used to detect E-cadherin and β- deoxycholate, and 0.1% sodium dodecyl sulfate). Protein cantenin were from Santa Cruz Biotechnology (Santa Cruz, concentrations of the lysates were determined with a Bio-Rad CA). The paraffin-embedded sections were pre-treated and protein assay kit (Hercules, CA). Immunoblotting analyses stained with antibodies by using the previously reported were performed as described previously [22]. Antibodies method [20,24]. The secondary antibodies against mouse or

PLOS ONE | www.plosone.org 3 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

rabbit IgG were supplied in an IHC kit (#CW2069) from Beijing Knockdown of RPS7 attenuates cellular apoptosis in CoWin Bioscience Co. Ltd (Beijing, China). ovarian cancer cells Statistical analysis. All data were analyzed by the Student We investigated the effect of RPS7 expression on cellular t test. P < 0.05 was considered statistically significant. apoptosis using annexin v fluorescence apoptosis assay. Although silencing of RPS7 expression slightly reduced cell Results apoptosis (Figure 2A-B), a decreased expression of the pro- apoptotic proteins p21waf1/cip1, p27cip/kip, BAX, BAK and BAD, and Knockdown of RPS7 increases cell proliferation an enhanced expression of the anti-apoptotic proteins Bcl-2 To investigate the function of RPS7 in human ovarian and Bcl-xl were detected in shRPS7-treated cells, compared cancer, we first detected the expression level of RPS7 in four with in control cells (Figure 1F and Figure 2C). These data human ovarian cancer cell lines and one immortalized human suggested that RPS7 might contribute to cellular apoptosis. ovarian surface epithelial cell line (T29). We found that the To validate the above results, we treated SKOV3-shRPS7, expression of RPS7 was higher in SKOV3, OVCA429 and OVCA429-shRNA, OVCA433-shRPS7 cells, and their control OVCA433 cells than in OVCA420 and T29 cells (Figure 1A). cells with cisplatin and used flow cytometry to detect cell Thus, we introduced retroviruses carrying shRNA against apoptosis indicated as the percentage of sub-G1 cells with PI RPS7 into SKOV3, OVCA429 and OVCA433 cells to generate staining (chromosomes < 2n). We found that, with the SKOV3-shRPS7, OVCA429-shRPS7 and OVCA433-shRPS7 cells (corresponding control cells were infected with stimulation of cisplatin at 2.5 µg/mL, the percentage of sub-G1 retroviruses expressing shGFP). As detected by Western cells was lower in SKOV3-shRPS7 (6.4%), OVCA429-shRNA blotting and realtime PCR, respectively, both the mRNA level (12.5%) or OVCA433-shRPS7 (8.5%) cells than that in and the protein level of RPS7 were decreased in cells treated SKOV3-shGFP (9.7%), OVCA429-shGFP (15.8%), or with RPS7 shRNA compared with in controls (Figure 1B-C). OVCA433-shGFP (13.7%) cells (Figure 2D-E) (P < 0.05). We next evaluated cell proliferation rates of the cell lines by However, without cisplatin treatment, the percentage of sub-G1 CCK8 assay. The results showed that cells expressing shRNA cells was much lower in both shRNA-treated cells and their had higher level of proliferation than the corresponding control control cells and had no statistical significance between cells expressing shGFP (Figure 1D). shRNA-treated cells and their control cells (Figure S1) (P > 0.05). These results suggested that knock down of RPS7 Down-regulation of RPS7 promotes cell cycle moderately attenuated cancer cells apoptosis and chemo- progression sensitivity to cisplatin treatment. Previous studies have shown that RPS7 may accelerate the G2-M phase transition of cell cycle during the early RPS7 regulates PI3K/AKT and MAPK signal networks development of zebrafish embryo [13]. We found that the cell Since PI3K/AKT and MAPK are two important intracellular population was decreased at the G0-G1 phase but increased at pathways involved in cell proliferation, cell cycle, and both S and G2-M phases in three cell lines treated with RPS7 apoptosis, we detected some important members in PI3K/AKT shRNA compared with in control cells (Figure 1E). To explore and MAPK signal networks by Western blotting. We found that the potential mechanism, we analyzed major proteins associated with cell cycle progression by Western blotting. The the levels of P85α and P110α, two main subunits of PI3K, were results in Figure 1F showed that p21Cip1/Waf1, an essential increased in SKOV3-shRPS7, OVCA429-shRPS7 or suppressor involved in the G1-S cell cycle transition, was OVCA433-shRPS7 cells (Figure 3A), compared within their remarkably decreased in SKOV3, OVCA429, and OVCA433 controls. AKT1 was decreased in SKOV3 and OVCA429, but cells after RPS7 was knocked down compared with in control not changed in OVCA433 after RPS7 was silenced (Figure cells (Figure 1F). Cyclin-dependant kinase 4 (Cdk 4) was 3A), whereas AKT2 was increased in all three cell lines in markedly increased in SKOV3-shRPS7 and OVCA429-shRPS7 which RSP7 was silenced (Figure 3A), compared with in cells cells, but was slightly increased in OVCA433-shRPS7 cells expressing shGFP. compared with in control cells expressing shGFP. Cyclin D1, a ERK 1/2, MEK1/2 and JNK1/2 were reduced in SKOV3- Cdk4 partner protein, was not altered in OVCA433-shRPS7, shRPS7 and OVCA429-shRPS7, but not in OVCA433-shRPS7 but was increased in OVCA429-shRPS7 and SKOV3-shRPS7 (Figure 3B), and the phosphorylation of MEK1/2 (Ser217/221), cells (Figure 1F), compared with in their control cells. The S- ERK1/2 ( Thr202/Tyr204), and JNK1/2(Thr183/Tyr185) was phase regulatory protein Cdk2 was markedly increased in all reduced in SKOV3-shRPS7, OVCA429-shRPS7 and cells whose expression of RPS7 was silenced, compared with OVCA433-shRPS7 cells, compared with in their controls in their control cells. Meanwhile, we found that cyclin B1, a G2- (Figure 3B). The phosphorylation of P38 (Thr180/Tyr182) was M transition-promoting protein, was increased in SKOV3, OVCA429 and OVCA433 cells after RPS7 was silenced. These reduced in all three ovarian cancer cell lines after RPS7 was data suggested that RPS7 promotes ovarian cancer cell cycle silenced, but P38 was not changed in SKOV3-shRPS7, progression in the G1-S and G2-M transitions, possibly through OVCA429-shRPS7 and OVCA433-shRPS7 cells (Figure 3B). suppression of p21Cip1/Waf1. Other factors, such as Cdk4 and Thus, based on the above results, we infer that RPS7 may cyclin B1, may also be involved in RPS7-associated cell cycle regulate cell proliferation, cell cycle, and cell apoptosis through regulation. PI3K/AKT and MAPK signaling.

PLOS ONE | www.plosone.org 4 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

Figure 1. Expression of RPS7 in ovarian cancer cells and immortalized ovarian surface epithelial cells and effects of RPS7 on cell proliferation and cell cycle. A, Detection of RPS7 expression by Western blotting in SKOV3, OVCA433, OVCA429, OVCA420 and T29 cells. B-C, Analyses of RPS7 by Western blotting (B) and realtime-PCR (C) in cells expressing shGFP or shRPS7. D, Detection of cell proliferation by CCK8 (P < 0.05). Error bars = 95% confidence intervals (CIs). E, Quantitative analysis of cell cycle distribution. Data from three independent experiments were analyzed (P < 0.05). Error bars = 95% CIs. F, Immunoblotting analysis of cell cycle regulatory proteins. β-actin was used as the loading control. doi: 10.1371/journal.pone.0079117.g001

PLOS ONE | www.plosone.org 5 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

Figure 2. Examination of cell apoptosis associated with RPS7. A, Apoptosis detected by flow cytometry. B, Quantitative analysis of apoptotic cells (P < 0.05). Error bars = 95% CIs. C, Immunoblotting analysis of apoptosis-associated proteins. β-actin was used as the loading control. D, Cisplatin-induced apoptosis indicated as sub-G1 population detected by flow cytometry. E, Quantitative analysis of sub-G1 cells (P < 0.05). Error bars = 95% CIs. doi: 10.1371/journal.pone.0079117.g002

PLOS ONE | www.plosone.org 6 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

shRPS7, but was not changed in OVCA433-shRPS7 cells (Figure 4E). MMP9 was increased in OVCA433-shRPS7 cells, but not in SKOV3-shRPS7 and OVCA429-shRPS7 cells (Figure 4E). MMP13 was increased in SKOV3-shRPS7 cells, but not in OVCA429-shRPS7 and OVCA433-shRPS7 cells (Figure 4E). Both E-cadherin and β-catenin were decreased in all ovarian cancer cell lines whose RPS7 was silenced (Figure 4E). Based on these results, RPS7 appears to influence ovarian cancer cell invasion and migration possibly through regulation of MMP2, E-cadherin, and β-catenin.

RPS7 is involved in ovarian tumorigenesis and metastasis To determine the roles of RPS7 in ovarian tumorigenesis, we first performed the anchorage-independent growth assay and found that SKOV3-shRPS7, OVCA429-shRPS7 and OVCA433-shRPS7 cells generated a greater number of colonies than their control cells (P < 0.05) (Figure 5A). To test the in vivo effect of RPS7 on tumorigenesis, SKOV3-shRPS7 or OVCA433-shRPS7 cells and their corresponding controls were subcutaneously injected into nude mice. As shown in Figure 5B-C, the tumors from nude mice injected with SKOV3- shRPS7 or OVCA433-shRPS7 cells grew faster and to a larger volume than those derived from control cells (P < 0.05). To further explore the function of RPS7 in cancer cell metastasis, SKOV3-shRPS7 or OVCA433-shRPS7 cells and their corresponding controls were intraperitoneally injected into nude mice. The results showed that silencing of RPS7 increased the number and weight of nodules (Figure 6B) and tumor ascites (Figure 6C-E) in mice injected with cells treated with shRPS7 compared with mice injected with cells expressing shGFP (P < 0.05). The analysis of tumors from nude mice by Figure 3. Detection of PI3K/AKT and MAPK signal immunohistochemical staining showed that the expression of pathways by Western blotting. A, Immunoblotting analysis both E-cadherin and β-catenin in tumor tissue from nude mice of molecules associated with PI3K/AKT signal pathway. β-actin injected with SKOV3-shRPS7 or OVCA433-shRPS7 cells were was used as the loading control. B, Immunoblotting analysis of reduced compared with those in tumor tissues derived from molecules associated with MAPK signal pathway. β-actin was mice injected with corresponding control cells (Figure 6A). used as the loading control. These results suggested that RSP7 might function to suppress doi: 10.1371/journal.pone.0079117.g003 tumor growth and metastasis during the development of ovarian cancer, which is consistent with the data that RSP7 Knockdown of RPS7 promotes cell invasion and inhibits cell invasion and migration (Figure 4). migration To investigate the function of RPS7 in invasion and Discussion migration, we performed assays using a high throughput transwell assay. We found that more SKOV3-shRPS7, In this study, we found that the aberrant expression of RPS7 OVCA429-shRPS7 and OVCA433-shRPS7 cells invaded might be associated with the development of ovarian cancer. through the membrane in the bottom chamber than their control Silencing of RPS7 promoted cell proliferation through cells (Figure 4A-B). We also detected migration speed by dysregulation of p21waf1/cip1, CDK2, CDK4, Cyclin B1, and Cyclin scratch assay and found that the migration speed of SKOV3- D1. Moreover, knockdown of RPS7 down-regulated the shRPS7, OVCA429-shRPS7 or OVCA433-shRPS7 cells was expression of pro-apoptotic factors such as p27cip/kip, BAX, BAK, increased after 24h culture compared with control cells (Figure and BAD, but up-regulated the expression of anti-apoptotic 3C-D). Further, we determined the expression of migration- factors, including Bcl-2 and Bcl-xl, which resulted in attenuated related proteins, including members of matrix apoptosis. Subsequently, we found that silencing of RPS7 metalloproteinase family (MMP), E-cadherin, and β-catenin. enhanced cell migration and invasion, which is consistent to Compared with in control cells, the expression level of MMP2 the regulatory functions of RPS7 in cell cycle and apoptosis. was up-regulated in SKOV3-shRPS7 cells and OVCA429- Further studies revealed that RPS7 might control the PI3K/AKT

PLOS ONE | www.plosone.org 7 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

Figure 4. Influence of cell invasion and migration by RPS7. A, Detection of cell invasion by using a high throughput screening multi-well insert 24-well two-chamber plates. B, Quantitative analysis of migration cells (P < 0.05). Error bars = 95% CIs. C, Quantitative analysis of migration speed with migration index (P < 0.05). Error bars = 95% CIs. D, Detection of migration speed by scratch assay. E, Immunoblotting analysis of metastasis-associated proteins. β-actin was used as the loading control. doi: 10.1371/journal.pone.0079117.g004

PLOS ONE | www.plosone.org 8 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

Figure 5. Xenograft tumor burden in mice with or without silencing of RPS7. A, In vitro tumorigenicity tested by soft agar assay (P < 0.05). Error bars = 95% CIs. B, In vivo tumorigenesis examined by animal assay. C, Subcutaneous tumor growth from mice injected with cells expressing either shGFP or shRPS7 (P < 0.05). Error bars = 95% CIs. doi: 10.1371/journal.pone.0079117.g005 and MAPK signal networks to inhibit ovarian tumorigenesis and G2/M arrest of cell cycle [5]. Further, they found that RPL13 metastasis. siRNA significantly enhanced the cellular sensitivity to certain Over the past decades, ribosomal proteins (RPs) have been DNA damaging agents and, concordantly, RPL13- found to play important roles in the development of many overexpressing cells displayed greater chemoresistance than cancers, including colon carcinoma, breast cancer, esophageal the parental cells, suggesting that an inverse correlation exists cancer, and hepatocellular carcinoma [25-27]. Extensive between RPL13 expression and chemosensitivity [5]. However, evidences on RPs’ tumorigenic functions have been Chien et al. reported that RPS19 in colorectal cancer (CRC) documented these years. Wu et al. found that overexpression possibly promoted cellular apoptosis through BAX/p53 of RPL6 in human gastric tissues predicted poor prognosis. Up- pathway, and that the level of fecal RPS19 might be used as a regulation of RPL6 accelerated cell growth and enhanced in vitro colony formation of gastric cancer cells, while down- favorable prognostic factor for CRC patients [29]. Thus, the regulation of RPL6 could suppress cell cycle progression, at function of RPs involved in cell proliferation, cell cycle, least partially through down-regulating cyclin E, indicating that apoptosis and tumorigenesis varies depending on tissues and RPL6 might be used as a novel therapeutic target for gastric cell types studied in literatures. Our results demonstrate that cancer [4]. Llanos et al. found that the ectopic overexpression RPS is abnormally expressed in ovarian cancer cell lines and of RPL37 attenuated the DNA damage response mediated by tissues and plays an important role in ovarian tumorigenesis p53, which supports the notion that the DNA damage-induced and chemo-sensitivity. However, the clinical significance of proteasome degradation of RPL37 constitutes a mechanistic RPS7 in ovarian cancer remains to be investigated in future. link between DNA damage and ribosomal stress pathway [28]. In conclusion, we demonstrate that RPS7 inhibits ovarian Kobayashi et al. confirmed that knockdown of RPL13 using tumor growth and metastasis through regulation of the small interfering RNA ( siRNA) resulted in drastic attenuation of PI3K/AKT and MAPK signal pathways. gastrointestinal cancer cell growth with significant G1 and

PLOS ONE | www.plosone.org 9 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

Figure 6. Analysis of intraperitoneally generated xenograft tumors and tissues. A, Dissection of xenograft tumor. B, Quantitative analysis of the number of nodules in nude mice (P < 0.05). Error bars = 95% CIs. C, Quantitative analysis of the weight of nodules in nude mice (P < 0.05). Error bars = 95% CIs. D, Quantitative analysis of the volume of tumor ascites in nude mice (P < 0.05). Error bars = 95% CIs. E, Immunohistochemical staining of E-cadherin and β-catenin in xenograft tumor tissues. Tissues were stained with mouse anti- E-cadherin andβ-catenin antibodies and visualized with donkey anti-mouse secondary antibody (Magnification×200). doi: 10.1371/journal.pone.0079117.g006

PLOS ONE | www.plosone.org 10 November 2013 | Volume 8 | Issue 11 | e79117 Ribosomal Protein S7 Regulates Ovarian Cancer

Supporting Information Author Contributions

Conceived and designed the experiments: GY XMG ZLW. Figure S1. Quantitative analysis of sub-G1 cells in cell Performed the experiments: ZLW JH LLL JM. Analyzed the lines without treatment of cisplatin (P > 0.05). data: ZHQ JMS HYG YHX JM. Contributed reagents/materials/ Error bars = 95% CIs. analysis tools: JM WG YW HZS. Wrote the manuscript: GY (PDF) ZLW XMG.

References

1. Morgan RJ, Alvarez RD, Armstrong DK, Burger RA, Castells M et al. stabilization of p53 protein, and activation of p53 function. Oncogene (2012) Ovarian Cancer, Version 3.2012. J Natl Compr Canc Netw 10: 26: 5029-5037. doi:10.1038/sj.onc.1210327. PubMed: 17310983. 1339-1349. 16. Auersperg N, Siemens CH, Myrdal SE (1984) Human ovarian surface 2. Seong KM, Jung SO, Kim HD, Kim HJ, Jung YJ et al. (2012) Yeast epithelium in primary culture. In Vitro 20: 743-755. doi:10.1007/ ribosomal protein S3 possesses a beta-lyase activity on damaged BF02618290. PubMed: 6083974. DNA. FEBS Lett 586: 356-361. doi:10.1016/j.febslet.2011.12.030. 17. Liu J, Yang G, Thompson-Lanza JA, Glassman A, Hayes K et al. PubMed: 22245673. (2004) A genetically defined model for human ovarian cancer. Cancer 3. Kobayashi Y, Kawakami K, Ohbayashi M, Kohyama N, Yamamoto T Res 64: 1655-1663. doi:10.1158/0008-5472.CAN-03-3380. PubMed: (2010) Ribosomal protein L3 mediated the transport of digoxin in 14996724. Xenopus laevis oocyte. J Toxicol Sci 35: 827-834. doi:10.2131/jts. 18. Riccardi C, Nicoletti I (2006) Analysis of apoptosis by propidium iodide 35.827. PubMed: 21139332. staining and flow cytometry. Nat Protoc ;1(3): 1458-1461. doi:10.1038/ 4. Wu Q, Gou Y, Wang Q, Jin H, Cui L et al. (2011) Down-regulation of nprot.2006.238. PubMed: 17406435. RPL6 by siRNA inhibits proliferation and cell cycle progression of 19. Yang G, Chang B, Yang F, Guo X, Cai KQ et al. (2010) Aurora kinase human gastric cancer cell lines. PLOS ONE 6: e26401. doi:10.1371/ A promotes ovarian tumorigenesis through dysregulation of the cell journal.pone.0026401. PubMed: 22043320. cycle and suppression of BRCA2. Clin Cancer Res 16: 3171-3181. doi: 5. Kobayashi T, Sasaki Y, Oshima Y, Yamamoto H, Mita H et al. (2006) 10.1158/1078-0432.CCR-09-3171. PubMed: 20423983. Activation of the ribosomal protein L13 gene in human gastrointestinal 20. Yang G, Rosen DG, Zhang Z, Bast RC Jr., Mills GB et al. (2006) The cancer. Int J Mol Med 18: 161-170. PubMed: 16786168. chemokine growth-regulated oncogene 1 (Gro-1) links RAS signaling to 6. Kondoh N, Schweinfest CW, Henderson KW, Papas TS (1992) the senescence of stromal fibroblasts and ovarian tumorigenesis. Proc Differential expression of S19 ribosomal protein, laminin-binding Natl Acad Sci U S A 103: 16472-16477. doi:10.1073/pnas.0605752103. protein, and human lymphocyte antigen class I messenger RNAs PubMed: 17060621. associated with colon carcinoma progression and differentiation. 21. Yang G, Rosen DG, Liu G, Yang F, Guo X et al. (2010) CXCR2 Cancer Res 52: 791-796. PubMed: 1339304. promotes ovarian cancer growth through dysregulated cell cycle, 7. Li C, Ge M, Yin Y, Luo M, Chen D (2012) Silencing expression of diminished apoptosis, and enhanced angiogenesis. Clin Cancer Res ribosomal protein L26 and L29 by RNA interfering inhibits proliferation 16: 3875-3886. doi:10.1158/1078-0432.CCR-10-0483. PubMed: of human pancreatic cancer PANC-1 cells. Mol Cell Biochem 370: 20505188. 127-139. doi:10.1007/s11010-012-1404-x. PubMed: 22868929. 22. Yang G, Thompson JA, Fang B, Liu J (2003) Silencing of H-ras gene 8. Bee A, Brewer D, Beesley C, Dodson A, Forootan S et al. (2011) expression by retrovirus-mediated siRNA decreases transformation siRNA knockdown of ribosomal protein gene RPL19 abrogates the efficiency and tumorgrowth in a model of human ovarian cancer. aggressive phenotype of human prostate cancer. PLOS ONE 6: Oncogene 22: 5694-5701. doi:10.1038/sj.onc.1206858. PubMed: e22672. doi:10.1371/journal.pone.0022672. PubMed: 21799931. 12944918. 9. Gou Y, Shi Y, Zhang Y, Nie Y, Wang J et al. (2010) Ribosomal protein 23. Liu G, Yang G, Chang B, Mercado-Uribe I, Huang M et al. (2010) L6 promotes growth and cell cycle progression through up-regulating Stanniocalcin 1 and ovarian tumorigenesis. J Natl Cancer Inst 102: cyclin E in gastric cancer cells. Biochem Biophys Res Commun 393: 812-827. doi:10.1093/jnci/djq127. PubMed: 20484106. 788-793. doi:10.1016/j.bbrc.2010.02.083. PubMed: 20171175. 24. Yang G, Xiao X, Rosen DG, Cheng X, Wu X et al. (2011) The biphasic 10. Wang Q, Yang C, Zhou J, Wang X, Wu M et al. (2001) Cloning and role of NF-kappaB in progression and chemoresistance of ovarian characterization of full-length human ribosomal protein L15 cDNA cancer. Clin Cancer Res 17: 2181-2194. doi: which was overexpressed in esophageal cancer. Gene 263: 205-209. 10.1158/1078-0432.CCR-10-3265. PubMed: 21339307. doi:10.1016/S0378-1119(00)00570-9. PubMed: 11223259. 25. Lindström MS, Nistér M (2010) Silencing of ribosomal protein S9 elicits 11. Wu LM, Wang SY, Wang SX, Huang YM, Li JG (2010) [Proliferation a multitude of cellular responses inhibiting the growth of cancer cells promotion and apoptotic inhibition effects of ribosomal protein RPL36A subsequent to p53 activation. PLOS ONE 5: e9578. doi:10.1371/ small interference RNA on U937 cells]. Zhongguo Shi Yan Xue Ye Xue journal.pone.0009578. PubMed: 20221446. Za Zhi 18: 344-349. PubMed: 20416165. 26. Guo X, Shi Y, Gou Y, Li J, Han S et al. (2011) Human ribosomal protein 12. Wu CT, Lin TY, Hsu HY, Sheu F, Ho CM et al. (2011) Ling Zhi-8 S13 promotes gastric cancer growth through down-regulating mediates p53-dependent growth arrest of lung cancer cells proliferation p27(Kip1). J Cell Mol Med 15: 296-306. doi:10.1111/j. via the ribosomal protein S7-MDM2-p53 pathway. Carcinogenesis 32: 1582-4934.2009.00969.x. PubMed: 19912438. 1890-1896. doi:10.1093/carcin/bgr221. PubMed: 21983128. 27. Bee A, Ke Y, Forootan S, Lin K, Beesley C et al. (2006) Ribosomal 13. Duan J, Ba Q, Wang Z, Hao M, Li X et al. (2011) Knockdown of protein l19 is a prognostic marker for human prostate cancer. Clin ribosomal protein S7 causes developmental abnormalities via p53 Cancer Res 12: 2061-2065. doi:10.1158/1078-0432.CCR-05-2445. dependent and independent pathways in zebrafish. Int J Biochem Cell PubMed: 16609016. Biol 43: 1218-1227. doi:10.1016/j.biocel.2011.04.015. PubMed: 28. Llanos S, Serrano M (2010) Depletion of ribosomal protein L37 occurs 21550419. in response to DNA damage and activates p53 through the L11/MDM2 14. Zhu Y, Poyurovsky MV, Li Y, Biderman L, Stahl J et al. (2009) pathway. Cell Cycle 9: 4005-4012. doi:10.4161/cc.9.19.13299. Ribosomal protein S7 is both a regulator and a substrate of MDM2. Mol PubMed: 20935493. Cell 35: 316-326. doi:10.1016/j.molcel.2009.07.014. PubMed: 29. Chien CC, Tu TC, Huang CJ, Yang SH, Lee CL (2012) Lowly 19683495. expressed ribosomal protein s19 in the feces of patients with colorectal 15. Chen D, Zhang Z, Li M, Wang W, Li Y et al. (2007) Ribosomal protein cancer. ISRN. Gastroenterologist 2012: 394545. S7 as a novel modulator of p53-MDM2 interaction: binding to MDM2,

PLOS ONE | www.plosone.org 11 November 2013 | Volume 8 | Issue 11 | e79117